IronPython.Runtime.Exceptions.ImportException:“No module named 'sklearn.linear_model._base'”

[abiaogo]

Environments:
Windows11, Visual Studio Version: 17.13.2 ;
IronPython version: 3.4.2； scikit-learn Version: 1.6.1 ;

Q1: Dose IronPython 3.4.2 support scikit-learn? is yes, how to solve this problem.
Q2: I want to load a line-regression model to c#, how to achieve this if IronPython 3.4.2 don't support scikit-learn.

[abiaogo]

Use Python 3.13.2 and scikit-learn 1.6.1 to generate the line-regression module and save it by "pickle.dump(model, file)"
Use Python 3.13.2 can load the model successfully by "model = pickle.load(file)"
The problem happened when I tried to load the model in .net8.0 use IronPython version: 3.4.2.

//C# code
using IronPython.Hosting;
class Program
{
static void Main(string[] args)
{
var engine = Python.CreateEngine();
engine.Execute(code);
dynamic py = engine.ExecuteFile("sklearn_line_load.py");
}
}

##sklearn_line_load.py
import pickle
with open('linear_model.pkl', 'rb') as file:
model = pickle.load(file)
def predict_price(data):
result = model.predict([data])
result = round(result[0])
return result

[Lamparter]

scikit-learn doesn't support Python 3.4.x. You will need to use 0.20 or earlier of scikit-learn to use it in IronPython.

[BCSharp]

scikit-learn depends on numpy. numpy is an extension package for Python and is incompatible with IronPython.

So scikit-learn (any version) does not work currently with IronPython 3. I suggest you look at Python for .NET whether it satisfies your needs.

[abiaogo]

thank you buddy. it's clear.